API referencePython APImodelversepipelinesAutomaticSpeechRecognitionPipelineAutomaticSpeechRecognitionPipeline pipe(audio) -> {"text": transcript}; a list of clips answers a list.